-----------------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  C:\Users\Wilfred\Dropbox\Chow&HanPapers\Race&Refugees\Replication Files\CH_jop_main.log
  log type:  text
 opened on:   8 Jul 2022, 11:23:21

. 
. set scheme plottig // set graphical style to plottig

. 
. // input main data
. use "CH_main.dta", clear

. 
. 
. 
. ********************************************************************************
. *** Figure 2 *******************************************************************
. ********************************************************************************
. 
. // generate mean values for outcome responses
. mean fair_b, over(outcome)

Mean estimation                   Number of obs   =      2,557

            1: outcome = 1
            2: outcome = 2

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
fair_b       |
           1 |   .4108949   .0137303      .3839713    .4378185
           2 |   .4622642   .0139848      .4348414    .4896869
--------------------------------------------------------------

.         estimates store m1

. mean trust_b, over(outcome)

Mean estimation                   Number of obs   =      2,557

            1: outcome = 1
            2: outcome = 2

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
trust_b      |
           1 |   .4357977   .0138381      .4086626    .4629327
           2 |   .5220126   .0140112      .4945381    .5494871
--------------------------------------------------------------

.         estimates store m2

. mean refugee_b, over(outcome)

Mean estimation                   Number of obs   =      2,557

            1: outcome = 1
            2: outcome = 2

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
refugee_b    |
           1 |   .4661479   .0139216      .4388491    .4934467
           2 |   .6438679   .0134317      .6175298    .6702061
--------------------------------------------------------------

.         estimates store m3

. mean american_b, over(outcome)

Mean estimation                   Number of obs   =      2,557

            1: outcome = 1
            2: outcome = 2

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
american_b   |
           1 |   .5019455   .0139535      .4745841    .5293069
           2 |   .6022013   .0137287      .5752807    .6291218
--------------------------------------------------------------

.         estimates store m4      

. 
. // generate plot for fairness of UNHCR 
. coefplot m1, ///
>                  vertical ciopts(recast(rcap) lcolor(black)) citop ///
>                  recast(bar) barwidth(.5) rescale(100) ///
>                  color(538g) ylabel(35(10)70) title("(a) UNHCR Is Fair") ///
>                  mlabel mlabposition(1) format(%9.3g) mlabcolor(black) ///
>                  xtitle("") ///
>                  xlabel(1 `""Non-Critical" "Report of the U.S.""' ///
>                                 2 `""Critical" "Report of the U.S.""', labcol(black)) ///
>                  ytitle("Mean Percent") ///
>                  addplot(scatter @b @at, ms(i) mlabel(@b) mlabpos(2) mlabcolor(black)) 

. graph save "g1.gph", replace            // save plot for use later
(note: file g1.gph not found)
(file g1.gph saved)

. 
. // generate plot for trustworthiness of UNHCR           
. coefplot m2, ///
>                  vertical ciopts(recast(rcap) lcolor(black)) citop ///
>                  recast(bar) barwidth(.5) rescale(100) ///
>                  color(538g) ///
>                  ylabel(35(10)70) ///
>                  title("(b) UNHCR Can Be Trusted") ///
>                  mlabel mlabposition(1) format(%9.3g) mlabcolor(black) ///
>                  xtitle("") ///
>                  xlabel(1 `""Non-Critical" "Report of the U.S.""' ///
>                                 2 `""Critical" "Report of the U.S.""', labcol(black)) ///
>                  ytitle("Mean Percent") ///
>                  addplot(scatter @b @at, ms(i) mlabel(@b) mlabpos(2) mlabcolor(black)) 

. graph save "g2.gph", replace    
(note: file g2.gph not found)
(file g2.gph saved)

. 
. // generate plot for outcome is good for refugees
. coefplot m3, ///
>                  vertical ciopts(recast(rcap) lcolor(black)) citop ///
>                  recast(bar) barwidth(.5) rescale(100) ///
>                  scheme(plottig) color(538g) ///
>                  ylabel(35(10)70) ///
>                  title("(c) Outcome Good for Refugees") ///
>                  mlabel mlabposition(1) format(%9.3g) mlabcolor(black) ///
>                  xtitle("") ///
>                  xlabel(1 `""Non-Critical" "Report of the U.S.""' ///
>                                 2 `""Critical" "Report of the U.S.""', labcol(black)) ///
>                  ytitle("Mean Percent") ///
>                  addplot(scatter @b @at, ms(i) mlabel(@b) mlabpos(2) mlabcolor(black)) 

. graph save "g3.gph", replace    
(note: file g3.gph not found)
(file g3.gph saved)

. 
. // generate plot for outcome is good for Americans
. coefplot m4, ///
>                  vertical ciopts(recast(rcap) lcolor(black)) citop ///
>                  recast(bar) barwidth(.5) rescale(100) ///
>                  scheme(plottig) color(538g) ///
>                  ylabel(35(10)70) ///
>                  title("(d) Outcome Good for Americans") ///
>                  mlabel mlabposition(1) format(%9.3g) mlabcolor(black) ///
>                  xtitle("") ///
>                  xlabel(1 `""Non-Critical" "Report of the U.S.""' ///
>                                 2 `""Critical" "Report of the U.S.""', labcol(black)) ///
>                  ytitle("Mean Percent") ///
>                  addplot(scatter @b @at, ms(i) mlabel(@b) mlabpos(2) mlabcolor(black)) 

. graph save "g4.gph", replace    
(note: file g4.gph not found)
(file g4.gph saved)

. 
. // combine saved .gph files into a single figure                        
. graph combine  "g1.gph" "g2.gph" "g3.gph" "g4.gph", ///
>                            xcommon ycommon cols(2) scale(1) ysize(4) xsize(5)      
(note:  clockdir by_legend_position not found in scheme, default attributes used)
(note:  clockdir by_legend_position not found in scheme, default attributes used)

. 
. ********************************************************************************
. *** Figure 3 *******************************************************************
. ********************************************************************************
. 
. // generate mean values for outcome responses by UNHCR panel treatement group
. mean fair_b, over(panel)

Mean estimation                   Number of obs   =      2,557

            1: panel = 1
            2: panel = 2
            3: panel = 3
            4: panel = 4

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
fair_b       |
           1 |   .3894081   .0192597      .3516419    .4271743
           2 |   .4188563   .0194115      .3807925      .45692
           3 |   .4150943   .0195537      .3767516    .4534371
           4 |   .5237342   .0198822      .4847472    .5627211
--------------------------------------------------------------

.         estimates store m1

. mean trust_b, over(panel)

Mean estimation                   Number of obs   =      2,557

            1: panel = 1
            2: panel = 2
            3: panel = 3
            4: panel = 4

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
trust_b      |
           1 |   .4252336   .0195268      .3869437    .4635235
           2 |   .4899536   .0196683      .4513863     .528521
           3 |   .4575472   .0197702      .4187798    .4963145
           4 |   .5427215   .0198319      .5038333    .5816097
--------------------------------------------------------------

.         estimates store m2

. mean refugee_b, over(panel)

Mean estimation                   Number of obs   =      2,557

            1: panel = 1
            2: panel = 2
            3: panel = 3
            4: panel = 4

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
refugee_b    |
           1 |    .546729   .0196624      .5081732    .5852848
           2 |   .5425039    .019601      .5040684    .5809394
           3 |   .5283019   .0198101      .4894564    .5671473
           4 |   .6012658   .0194922      .5630438    .6394879
--------------------------------------------------------------

.         estimates store m3

. mean american_b, over(panel)

Mean estimation                   Number of obs   =      2,557

            1: panel = 1
            2: panel = 2
            3: panel = 3
            4: panel = 4

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
american_b   |
           1 |   .5202492   .0197326      .4815557    .5589428
           2 |   .5440495   .0195957      .5056243    .5824746
           3 |   .5393082   .0197805      .5005208    .5780956
           4 |   .6044304   .0194657      .5662602    .6426005
--------------------------------------------------------------

.         estimates store m4

.         
. // generate plot for procedural outcome measures
. coefplot m1, bylabel("(a) UNHCR is Fair") || ///
>                  m2, bylabel("(b) UNHCR can be Trusted") ///
>                  vertical ciopts(recast(rcap) lcolor(black)) citop ///
>                  recast(bar) barwidth(.5) rescale(100) ///
>                  scheme(plottig) color(538g) ///
>                  mlabel mlabposition(1) format(%9.3g) mlabcolor(black) ///
>                  xtitle("Gender and Racial Distribution of Panel") ///
>                  xlabel(1 `""All-White" "All-Male""' ///
>                                 2 `""Mixed-Race" "All-Male""' ///
>                                 3 `""All-White" "Mixed-Gender""' ///
>                                 4 `""Mixed-Race" "Mixed-Gender""', labcol(black)) ///
>                  ytitle("Mean Percent") ///
>                  title("Procedural Legitimacy") ///
>                  addplot(scatter @b @at, ms(i) mlabel(@b) mlabpos(2) mlabcolor(black)) 
(note:  clockdir by_legend_position not found in scheme, default attributes used)

. graph save "g5.gph", replace    
(note: file g5.gph not found)
(file g5.gph saved)

. 
. // generate plot for substantive outcome measures
. coefplot m3, bylabel("(c) Outcome Good for Refugees") || ///
>                  m4, bylabel("(d) Outcome Good for Americans") ///
>                  vertical ciopts(recast(rcap) lcolor(black)) citop ///
>                  recast(bar) barwidth(.5) rescale(100) ///
>                  scheme(plottig) color(538g) ///
>                  mlabel mlabposition(1) format(%9.3g) mlabcolor(black) ///
>                  xtitle("Gender and Racial Distribution of Panel") ///
>                  xlabel(1 `""All-White" "All-Male""' ///
>                                 2 `""Mixed-Race" "All-Male""' ///
>                                 3 `""All-White" "Mixed-Gender""' ///
>                                 4 `""Mixed-Race" "Mixed-Gender""', labcol(black)) ///
>                  ytitle("Mean Percent") ///
>                  title("Substantive Legitimacy") ///
>                  addplot(scatter @b @at, ms(i) mlabel(@b) mlabpos(2) mlabcolor(black)) 
(note:  clockdir by_legend_position not found in scheme, default attributes used)

. graph save "g6.gph", replace    
(note: file g6.gph not found)
(file g6.gph saved)

. 
. // combine saved .gph files into a single figure
. graph combine "g5.gph" "g6.gph", xcommon ycommon cols(1) scale(1.25)
(note:  clockdir by_legend_position not found in scheme, default attributes used)
(note:  clockdir by_legend_position not found in scheme, default attributes used)
(note:  clockdir by_legend_position not found in scheme, default attributes used)

. 
.           
. ********************************************************************************
. *** Figure 4 *******************************************************************
. ********************************************************************************
. 
. // generate mean values for procedural outcome by UNHCR panel treatment group
. mean fair_b if outcome==1, over(panel)

Mean estimation                   Number of obs   =      1,285

            1: panel = 1
            2: panel = 2
            3: panel = 3
            4: panel = 4

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
fair_b       |
           1 |   .3567073   .0264903      .3047383    .4086763
           2 |   .4006211   .0273505      .3469646    .4542777
           3 |    .364486   .0269047       .311704     .417268
           4 |   .5254777    .028225      .4701056    .5808498
--------------------------------------------------------------

.         estimates store f0

. mean fair_b if outcome==2, over(panel)

Mean estimation                   Number of obs   =      1,272

            1: panel = 1
            2: panel = 2
            3: panel = 3
            4: panel = 4

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
fair_b       |
           1 |   .4235669   .0279295      .3687739    .4783599
           2 |   .4369231   .0275559      .3828631     .490983
           3 |   .4666667   .0281539      .4114335    .5218998
           4 |   .5220126   .0280556      .4669723    .5770529
--------------------------------------------------------------

.         estimates store f1

. mean trust_b if outcome==1, over(panel)

Mean estimation                   Number of obs   =      1,285

            1: panel = 1
            2: panel = 2
            3: panel = 3
            4: panel = 4

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
trust_b      |
           1 |   .3810976   .0268569      .3284094    .4337857
           2 |   .4751553   .0278728       .420474    .5298365
           3 |   .3925234   .0272975      .3389708    .4460759
           4 |   .4968153   .0282611      .4413723    .5522583
--------------------------------------------------------------

.         estimates store t0

. mean trust_b if outcome==2, over(panel)

Mean estimation                   Number of obs   =      1,272

            1: panel = 1
            2: panel = 2
            3: panel = 3
            4: panel = 4

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
trust_b      |
           1 |   .4713376   .0282152      .4159841    .5266911
           2 |   .5046154   .0277766      .4501224    .5591084
           3 |   .5238095   .0281846       .468516     .579103
           4 |   .5880503   .0276439      .5338176    .6422831
--------------------------------------------------------------

.         estimates store t1

. 
. // generate plot for procedural outcome measure (unhcr is fair)
. coefplot f0, bylabel("(a) Non-Critical Report of the U.S.") || ///
>                  f1, bylabel("(b) Critical Report of the U.S.") ///
>                  vertical ciopts(recast(rcap) lcolor(black)) citop ///
>                  recast(bar) barwidth(.5) rescale(100) ///
>                  scheme(plottig) color(538g) ///
>                  mlabel mlabposition(1) format(%9.3g) mlabcolor(black) ///
>                  xtitle("Gender and Racial Distribution of Panel") ///
>                  xlabel(1 `""All-White" "All-Male""' ///
>                                 2 `""Mixed-Race" "All-Male""' ///
>                                 3 `""All-White" "Mixed-Gender""' ///
>                                 4 `""Mixed-Race" "Mixed-Gender""', labcol(black)) ///
>                  ytitle("Mean Percent") ///
>                  title("UNHCR is Fair") ///
>                  addplot(scatter @b @at, ms(i) mlabel(@b) mlabpos(2) mlabcolor(black))
(note:  clockdir by_legend_position not found in scheme, default attributes used)

. graph save "g7.gph", replace                     
(note: file g7.gph not found)
(file g7.gph saved)

. 
. // generate plot for procedural outcome measure (unhcr is trustworthy)
. coefplot t0, bylabel("(c) Non-Critical Report of the U.S.") || ///
>                  t1, bylabel("(d) Critical Report of the U.S.") ///
>                  vertical ciopts(recast(rcap) lcolor(black)) citop ///
>                  recast(bar) barwidth(.5) rescale(100) ///
>                  scheme(plottig) color(538g) ///
>                  mlabel mlabposition(1) format(%9.3g) mlabcolor(black) ///
>                  xtitle("Gender and Racial Distribution of Panel") ///
>                  xlabel(1 `""All-White" "All-Male""' ///
>                                 2 `""Mixed-Race" "All-Male""' ///
>                                 3 `""All-White" "Mixed-Gender""' ///
>                                 4 `""Mixed-Race" "Mixed-Gender""', labcol(black)) ///
>                  ytitle("Mean Percent") ///
>                  title("UNHCR can be Trusted") ///
>                  addplot(scatter @b @at, ms(i) mlabel(@b) mlabpos(2) mlabcolor(black))
(note:  clockdir by_legend_position not found in scheme, default attributes used)

. graph save "g8.gph", replace                     
(note: file g8.gph not found)
(file g8.gph saved)

. 
. // combine saved .gph files into a single figure
. graph combine "g7.gph" "g8.gph", xcommon  cols(1) scale(1.25)   
(note:  clockdir by_legend_position not found in scheme, default attributes used)
(note:  clockdir by_legend_position not found in scheme, default attributes used)
(note:  clockdir by_legend_position not found in scheme, default attributes used)

. 
. ********************************************************************************
. *** Figure 5 *******************************************************************
. ********************************************************************************
. 
. mean refugee_b if outcome==1, over(panel)

Mean estimation                   Number of obs   =      1,285

            1: panel = 1
            2: panel = 2
            3: panel = 3
            4: panel = 4

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
refugee_b    |
           1 |   .4420732   .0274639      .3881942    .4959522
           2 |   .4689441   .0278534      .4143009    .5235873
           3 |   .4361371   .0277219      .3817518    .4905223
           4 |   .5191083    .028241      .4637047    .5745119
--------------------------------------------------------------

.         estimates store r0

. mean refugee_b if outcome==2, over(panel)

Mean estimation                   Number of obs   =      1,272

            1: panel = 1
            2: panel = 2
            3: panel = 3
            4: panel = 4

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
refugee_b    |
           1 |    .656051     .02685      .6033758    .7087261
           2 |   .6153846    .027028      .5623602     .668409
           3 |   .6222222   .0273606      .5685453    .6758992
           4 |   .6823899   .0261477      .6310925    .7336874
--------------------------------------------------------------

.         estimates store r1

. mean american_b if outcome==1, over(panel)

Mean estimation                   Number of obs   =      1,285

            1: panel = 1
            2: panel = 2
            3: panel = 3
            4: panel = 4

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
american_b   |
           1 |   .4420732   .0274639      .3881942    .4959522
           2 |   .5062112   .0279051      .4514665    .5609558
           3 |   .4890966   .0279442      .4342753    .5439179
           4 |   .5732484   .0279568      .5184025    .6280943
--------------------------------------------------------------

.         estimates store a0

. mean american_b if outcome==2, over(panel)

Mean estimation                   Number of obs   =      1,272

            1: panel = 1
            2: panel = 2
            3: panel = 3
            4: panel = 4

--------------------------------------------------------------
        Over |       Mean   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
american_b   |
           1 |   .6019108   .0276684      .5476301    .6561916
           2 |   .5815385   .0274059      .5277726    .6353043
           3 |   .5904762   .0277508      .5360337    .6449187
           4 |   .6352201   .0270363      .5821794    .6882609
--------------------------------------------------------------

.         estimates store a1

.         
. // generate plot for substative outcome measure (refugee outcome)       
. coefplot r0, bylabel("(a) Non-Critical Report of the U.S.") || ///
>                  r1, bylabel("(b) Critical Report of the U.S.") ///
>                  vertical ciopts(recast(rcap) lcolor(black)) citop ///
>                  recast(bar) barwidth(.5) rescale(100) ///
>                  scheme(plottig) color(538g) ///
>                  mlabel mlabposition(1) format(%9.3g) mlabcolor(black) ///
>                  xtitle("Gender and Racial Distribution of Panel") ///
>                  xlabel(1 `""All-White" "All-Male""' ///
>                                 2 `""Mixed-Race" "All-Male""' ///
>                                 3 `""All-White" "Mixed-Gender""' ///
>                                 4 `""Mixed-Race" "Mixed-Gender""', labcol(black)) ///
>                  ytitle("Mean Percent") ///
>                  title("Outcome Good for Refugees") ///
>                  addplot(scatter @b @at, ms(i) mlabel(@b) mlabpos(2) mlabcolor(black))
(note:  clockdir by_legend_position not found in scheme, default attributes used)

. graph save "g9.gph", replace                                     
(note: file g9.gph not found)
(file g9.gph saved)

. 
. // generate plot for procedural outcome measure (american outcome)               
. coefplot a0, bylabel("(c) Non-Critical Report of the U.S.") || ///
>                  a1, bylabel("(d) Critical Report of the U.S.") ///
>                  vertical ciopts(recast(rcap) lcolor(black)) citop ///
>                  recast(bar) barwidth(.5)  rescale(100) ///
>                  scheme(plottig) color(538g) ///
>                  mlabel mlabposition(1) format(%9.3g) mlabcolor(black) ///
>                  xtitle("Gender and Racial Distribution of Panel") ///
>                  xlabel(1 `""All-White" "All-Male""' ///
>                                 2 `""Mixed-Race" "All-Male""' ///
>                                 3 `""All-White" "Mixed-Gender""' ///
>                                 4 `""Mixed-Race" "Mixed-Gender""', labcol(black)) ///
>                  ytitle("Mean Percent") ///
>                  title("Outcome Good for Americans") ///
>                  addplot(scatter @b @at, ms(i) mlabel(@b) mlabpos(2) mlabcolor(black))           
(note:  clockdir by_legend_position not found in scheme, default attributes used)

. graph save "g10.gph", replace   
(note: file g10.gph not found)
(file g10.gph saved)

. 
. // combine saved .gph files into a single figure
. graph combine "g9.gph" "g10.gph", xcommon  cols(1) scale(1.25)          
(note:  clockdir by_legend_position not found in scheme, default attributes used)
(note:  clockdir by_legend_position not found in scheme, default attributes used)
(note:  clockdir by_legend_position not found in scheme, default attributes used)

. 
. 
. ********************************************************************************
. *** Figure 6 *******************************************************************
. ********************************************************************************
. 
. 
. // generate estimates for outcome for american variable
. probit american_b i.outcome##i.party, robust

Iteration 0:   log pseudolikelihood =  -1695.569  
Iteration 1:   log pseudolikelihood = -1555.0179  
Iteration 2:   log pseudolikelihood = -1554.5332  
Iteration 3:   log pseudolikelihood = -1554.5332  

Probit regression                               Number of obs     =      2,466
                                                Wald chi2(5)      =     268.55
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1554.5332               Pseudo R2         =     0.0832

-------------------------------------------------------------------------------
              |               Robust
   american_b |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    2.outcome |   .2621057   .0882951     2.97   0.003     .0890506    .4351608
              |
        party |
           2  |  -.2488747   .0853137    -2.92   0.004    -.4160864    -.081663
           3  |   .9048067    .095935     9.43   0.000     .7167775    1.092836
              |
outcome#party |
         2 2  |   .8408706   .1239748     6.78   0.000     .5978845    1.083857
         2 3  |  -1.154231   .1330075    -8.68   0.000    -1.414921   -.8935409
              |
        _cons |  -.1390486   .0610211    -2.28   0.023    -.2586477   -.0194494
-------------------------------------------------------------------------------

.         margins i.outcome, at(party=1) vsquish post

Adjusted predictions                            Number of obs     =      2,466
Model VCE    : Robust

Expression   : Pr(american_b), predict()
at           : party           =           1

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     outcome |
          1  |   .4447059   .0241097    18.45   0.000     .3974518      .49196
          2  |   .5489691   .0252667    21.73   0.000     .4994472     .598491
------------------------------------------------------------------------------

.         estimates store ind1

. probit american_b i.outcome##i.party, robust

Iteration 0:   log pseudolikelihood =  -1695.569  
Iteration 1:   log pseudolikelihood = -1555.0179  
Iteration 2:   log pseudolikelihood = -1554.5332  
Iteration 3:   log pseudolikelihood = -1554.5332  

Probit regression                               Number of obs     =      2,466
                                                Wald chi2(5)      =     268.55
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1554.5332               Pseudo R2         =     0.0832

-------------------------------------------------------------------------------
              |               Robust
   american_b |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    2.outcome |   .2621057   .0882951     2.97   0.003     .0890506    .4351608
              |
        party |
           2  |  -.2488747   .0853137    -2.92   0.004    -.4160864    -.081663
           3  |   .9048067    .095935     9.43   0.000     .7167775    1.092836
              |
outcome#party |
         2 2  |   .8408706   .1239748     6.78   0.000     .5978845    1.083857
         2 3  |  -1.154231   .1330075    -8.68   0.000    -1.414921   -.8935409
              |
        _cons |  -.1390486   .0610211    -2.28   0.023    -.2586477   -.0194494
-------------------------------------------------------------------------------

.         margins i.outcome, at(party=2) vsquish post

Adjusted predictions                            Number of obs     =      2,466
Model VCE    : Robust

Expression   : Pr(american_b), predict()
at           : party           =           2

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     outcome |
          1  |   .3490364   .0220619    15.82   0.000     .3057958     .392277
          2  |   .7627119   .0195855    38.94   0.000     .7243249    .8010988
------------------------------------------------------------------------------

.         estimates store dem1

. probit american_b i.outcome##i.party, robust

Iteration 0:   log pseudolikelihood =  -1695.569  
Iteration 1:   log pseudolikelihood = -1555.0179  
Iteration 2:   log pseudolikelihood = -1554.5332  
Iteration 3:   log pseudolikelihood = -1554.5332  

Probit regression                               Number of obs     =      2,466
                                                Wald chi2(5)      =     268.55
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1554.5332               Pseudo R2         =     0.0832

-------------------------------------------------------------------------------
              |               Robust
   american_b |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    2.outcome |   .2621057   .0882951     2.97   0.003     .0890506    .4351608
              |
        party |
           2  |  -.2488747   .0853137    -2.92   0.004    -.4160864    -.081663
           3  |   .9048067    .095935     9.43   0.000     .7167775    1.092836
              |
outcome#party |
         2 2  |   .8408706   .1239748     6.78   0.000     .5978845    1.083857
         2 3  |  -1.154231   .1330075    -8.68   0.000    -1.414921   -.8935409
              |
        _cons |  -.1390486   .0610211    -2.28   0.023    -.2586477   -.0194494
-------------------------------------------------------------------------------

.         margins i.outcome, at(party=3) vsquish post

Adjusted predictions                            Number of obs     =      2,466
Model VCE    : Robust

Expression   : Pr(american_b), predict()
at           : party           =           3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     outcome |
          1  |   .7780899   .0220276    35.32   0.000     .7349166    .8212631
          2  |   .4497207   .0262972    17.10   0.000     .3981791    .5012622
------------------------------------------------------------------------------

.         estimates store rep1    

. // generate estimates for outcome for refugee variable
. probit refugee_b i.outcome##i.party, robust

Iteration 0:   log pseudolikelihood = -1694.4903  
Iteration 1:   log pseudolikelihood =  -1562.131  
Iteration 2:   log pseudolikelihood = -1561.8006  
Iteration 3:   log pseudolikelihood = -1561.8006  

Probit regression                               Number of obs     =      2,466
                                                Wald chi2(5)      =     254.33
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1561.8006               Pseudo R2         =     0.0783

-------------------------------------------------------------------------------
              |               Robust
    refugee_b |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    2.outcome |   .4468985   .0888418     5.03   0.000     .2727719    .6210252
              |
        party |
           2  |  -.2654635   .0859761    -3.09   0.002    -.4339736   -.0969535
           3  |   .7624739   .0933273     8.17   0.000     .5795558    .9453921
              |
outcome#party |
         2 2  |   .7393773   .1247693     5.93   0.000     .4948339    .9839207
         2 3  |  -.9193389   .1313565    -7.00   0.000    -1.176793    -.661885
              |
        _cons |  -.1988847    .061246    -3.25   0.001    -.3189247   -.0788447
-------------------------------------------------------------------------------

.         margins i.outcome, at(party=1) vsquish post

Adjusted predictions                            Number of obs     =      2,466
Model VCE    : Robust

Expression   : Pr(refugee_b), predict()
at           : party           =           1

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     outcome |
          1  |   .4211765   .0239551    17.58   0.000     .3742253    .4681277
          2  |   .5979381    .024897    24.02   0.000     .5491409    .6467353
------------------------------------------------------------------------------

.         estimates store ind2

. probit refugee_b i.outcome##i.party, robust

Iteration 0:   log pseudolikelihood = -1694.4903  
Iteration 1:   log pseudolikelihood =  -1562.131  
Iteration 2:   log pseudolikelihood = -1561.8006  
Iteration 3:   log pseudolikelihood = -1561.8006  

Probit regression                               Number of obs     =      2,466
                                                Wald chi2(5)      =     254.33
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1561.8006               Pseudo R2         =     0.0783

-------------------------------------------------------------------------------
              |               Robust
    refugee_b |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    2.outcome |   .4468985   .0888418     5.03   0.000     .2727719    .6210252
              |
        party |
           2  |  -.2654635   .0859761    -3.09   0.002    -.4339736   -.0969535
           3  |   .7624739   .0933273     8.17   0.000     .5795558    .9453921
              |
outcome#party |
         2 2  |   .7393773   .1247693     5.93   0.000     .4948339    .9839207
         2 3  |  -.9193389   .1313565    -7.00   0.000    -1.176793    -.661885
              |
        _cons |  -.1988847    .061246    -3.25   0.001    -.3189247   -.0788447
-------------------------------------------------------------------------------

.         margins i.outcome, at(party=2) vsquish post

Adjusted predictions                            Number of obs     =      2,466
Model VCE    : Robust

Expression   : Pr(refugee_b), predict()
at           : party           =           2

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     outcome |
          1  |   .3211991   .0216117    14.86   0.000     .2788411    .3635572
          2  |   .7648305    .019525    39.17   0.000     .7265623    .8030987
------------------------------------------------------------------------------

.         estimates store dem2

. probit refugee_b i.outcome##i.party, robust

Iteration 0:   log pseudolikelihood = -1694.4903  
Iteration 1:   log pseudolikelihood =  -1562.131  
Iteration 2:   log pseudolikelihood = -1561.8006  
Iteration 3:   log pseudolikelihood = -1561.8006  

Probit regression                               Number of obs     =      2,466
                                                Wald chi2(5)      =     254.33
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1561.8006               Pseudo R2         =     0.0783

-------------------------------------------------------------------------------
              |               Robust
    refugee_b |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    2.outcome |   .4468985   .0888418     5.03   0.000     .2727719    .6210252
              |
        party |
           2  |  -.2654635   .0859761    -3.09   0.002    -.4339736   -.0969535
           3  |   .7624739   .0933273     8.17   0.000     .5795558    .9453921
              |
outcome#party |
         2 2  |   .7393773   .1247693     5.93   0.000     .4948339    .9839207
         2 3  |  -.9193389   .1313565    -7.00   0.000    -1.176793    -.661885
              |
        _cons |  -.1988847    .061246    -3.25   0.001    -.3189247   -.0788447
-------------------------------------------------------------------------------

.         margins i.outcome, at(party=3) vsquish post

Adjusted predictions                            Number of obs     =      2,466
Model VCE    : Robust

Expression   : Pr(refugee_b), predict()
at           : party           =           3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     outcome |
          1  |   .7134831   .0239679    29.77   0.000     .6665069    .7604594
          2  |   .5363128   .0263614    20.34   0.000     .4846455    .5879802
------------------------------------------------------------------------------

.         estimates store rep2    

. // generate estimates for outcome for fairness variable
. probit fair_b i.outcome##i.party, robust

Iteration 0:   log pseudolikelihood = -1690.5152  
Iteration 1:   log pseudolikelihood =  -1603.003  
Iteration 2:   log pseudolikelihood = -1602.9164  
Iteration 3:   log pseudolikelihood = -1602.9164  

Probit regression                               Number of obs     =      2,466
                                                Wald chi2(5)      =     170.74
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1602.9164               Pseudo R2         =     0.0518

-------------------------------------------------------------------------------
              |               Robust
       fair_b |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    2.outcome |   .0774224   .0896184     0.86   0.388    -.0982265    .2530713
              |
        party |
           2  |  -.2418976   .0877405    -2.76   0.006    -.4138658   -.0699295
           3  |   .7496596    .092437     8.11   0.000     .5684865    .9308328
              |
outcome#party |
         2 2  |   .7198462   .1234546     5.83   0.000     .4778797    .9618127
         2 3  |  -.7587412   .1312544    -5.78   0.000    -1.015995   -.5014874
              |
        _cons |  -.3521765   .0621974    -5.66   0.000    -.4740811   -.2302719
-------------------------------------------------------------------------------

.         margins i.outcome, at(party=1) vsquish post

Adjusted predictions                            Number of obs     =      2,466
Model VCE    : Robust

Expression   : Pr(fair_b), predict()
at           : party           =           1

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     outcome |
          1  |   .3623529   .0233211    15.54   0.000     .3166444    .4080615
          2  |   .3917526   .0247867    15.80   0.000     .3431716    .4403336
------------------------------------------------------------------------------

.         estimates store ind3

. probit fair_b i.outcome##i.party, robust

Iteration 0:   log pseudolikelihood = -1690.5152  
Iteration 1:   log pseudolikelihood =  -1603.003  
Iteration 2:   log pseudolikelihood = -1602.9164  
Iteration 3:   log pseudolikelihood = -1602.9164  

Probit regression                               Number of obs     =      2,466
                                                Wald chi2(5)      =     170.74
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1602.9164               Pseudo R2         =     0.0518

-------------------------------------------------------------------------------
              |               Robust
       fair_b |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    2.outcome |   .0774224   .0896184     0.86   0.388    -.0982265    .2530713
              |
        party |
           2  |  -.2418976   .0877405    -2.76   0.006    -.4138658   -.0699295
           3  |   .7496596    .092437     8.11   0.000     .5684865    .9308328
              |
outcome#party |
         2 2  |   .7198462   .1234546     5.83   0.000     .4778797    .9618127
         2 3  |  -.7587412   .1312544    -5.78   0.000    -1.015995   -.5014874
              |
        _cons |  -.3521765   .0621974    -5.66   0.000    -.4740811   -.2302719
-------------------------------------------------------------------------------

.         margins i.outcome, at(party=2) vsquish post

Adjusted predictions                            Number of obs     =      2,466
Model VCE    : Robust

Expression   : Pr(fair_b), predict()
at           : party           =           2

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     outcome |
          1  |   .2762313    .020695    13.35   0.000     .2356698    .3167928
          2  |   .5805085   .0227187    25.55   0.000     .5359807    .6250363
------------------------------------------------------------------------------

.         estimates store dem3

. probit fair_b i.outcome##i.party, robust

Iteration 0:   log pseudolikelihood = -1690.5152  
Iteration 1:   log pseudolikelihood =  -1603.003  
Iteration 2:   log pseudolikelihood = -1602.9164  
Iteration 3:   log pseudolikelihood = -1602.9164  

Probit regression                               Number of obs     =      2,466
                                                Wald chi2(5)      =     170.74
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1602.9164               Pseudo R2         =     0.0518

-------------------------------------------------------------------------------
              |               Robust
       fair_b |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    2.outcome |   .0774224   .0896184     0.86   0.388    -.0982265    .2530713
              |
        party |
           2  |  -.2418976   .0877405    -2.76   0.006    -.4138658   -.0699295
           3  |   .7496596    .092437     8.11   0.000     .5684865    .9308328
              |
outcome#party |
         2 2  |   .7198462   .1234546     5.83   0.000     .4778797    .9618127
         2 3  |  -.7587412   .1312544    -5.78   0.000    -1.015995   -.5014874
              |
        _cons |  -.3521765   .0621974    -5.66   0.000    -.4740811   -.2302719
-------------------------------------------------------------------------------

.         margins i.outcome, at(party=3) vsquish post

Adjusted predictions                            Number of obs     =      2,466
Model VCE    : Robust

Expression   : Pr(fair_b), predict()
at           : party           =           3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     outcome |
          1  |   .6544944   .0252083    25.96   0.000      .605087    .7039017
          2  |   .3882682   .0257628    15.07   0.000      .337774    .4387623
------------------------------------------------------------------------------

.         estimates store rep3            

. // generate estimates for outcome for trust variable
. probit trust_b i.outcome##i.party, robust

Iteration 0:   log pseudolikelihood = -1707.0221  
Iteration 1:   log pseudolikelihood = -1619.8159  
Iteration 2:   log pseudolikelihood = -1619.7296  
Iteration 3:   log pseudolikelihood = -1619.7296  

Probit regression                               Number of obs     =      2,466
                                                Wald chi2(5)      =     169.98
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1619.7296               Pseudo R2         =     0.0511

-------------------------------------------------------------------------------
              |               Robust
      trust_b |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    2.outcome |   .1790028   .0888981     2.01   0.044     .0047657    .3532399
              |
        party |
           2  |  -.1919573    .086694    -2.21   0.027    -.3618744   -.0220402
           3  |   .7756426   .0927817     8.36   0.000     .5937937    .9574915
              |
outcome#party |
         2 2  |    .640166   .1226601     5.22   0.000     .3997565    .8805754
         2 3  |  -.7512935   .1307307    -5.75   0.000    -1.007521    -.495066
              |
        _cons |  -.3085721   .0618708    -4.99   0.000    -.4298366   -.1873076
-------------------------------------------------------------------------------

.         margins i.outcome, at(party=1) vsquish post

Adjusted predictions                            Number of obs     =      2,466
Model VCE    : Robust

Expression   : Pr(trust_b), predict()
at           : party           =           1

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     outcome |
          1  |   .3788235   .0235353    16.10   0.000     .3326952    .4249519
          2  |   .4484536   .0252535    17.76   0.000     .3989576    .4979496
------------------------------------------------------------------------------

.         estimates store ind4

. probit trust_b i.outcome##i.party, robust

Iteration 0:   log pseudolikelihood = -1707.0221  
Iteration 1:   log pseudolikelihood = -1619.8159  
Iteration 2:   log pseudolikelihood = -1619.7296  
Iteration 3:   log pseudolikelihood = -1619.7296  

Probit regression                               Number of obs     =      2,466
                                                Wald chi2(5)      =     169.98
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1619.7296               Pseudo R2         =     0.0511

-------------------------------------------------------------------------------
              |               Robust
      trust_b |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    2.outcome |   .1790028   .0888981     2.01   0.044     .0047657    .3532399
              |
        party |
           2  |  -.1919573    .086694    -2.21   0.027    -.3618744   -.0220402
           3  |   .7756426   .0927817     8.36   0.000     .5937937    .9574915
              |
outcome#party |
         2 2  |    .640166   .1226601     5.22   0.000     .3997565    .8805754
         2 3  |  -.7512935   .1307307    -5.75   0.000    -1.007521    -.495066
              |
        _cons |  -.3085721   .0618708    -4.99   0.000    -.4298366   -.1873076
-------------------------------------------------------------------------------

.         margins i.outcome, at(party=2) vsquish post

Adjusted predictions                            Number of obs     =      2,466
Model VCE    : Robust

Expression   : Pr(trust_b), predict()
at           : party           =           2

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     outcome |
          1  |   .3083512   .0213745    14.43   0.000      .266458    .3502444
          2  |       .625   .0222881    28.04   0.000     .5813162    .6686838
------------------------------------------------------------------------------

.         estimates store dem4

. probit trust_b i.outcome##i.party, robust

Iteration 0:   log pseudolikelihood = -1707.0221  
Iteration 1:   log pseudolikelihood = -1619.8159  
Iteration 2:   log pseudolikelihood = -1619.7296  
Iteration 3:   log pseudolikelihood = -1619.7296  

Probit regression                               Number of obs     =      2,466
                                                Wald chi2(5)      =     169.98
                                                Prob > chi2       =     0.0000
Log pseudolikelihood = -1619.7296               Pseudo R2         =     0.0511

-------------------------------------------------------------------------------
              |               Robust
      trust_b |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    2.outcome |   .1790028   .0888981     2.01   0.044     .0047657    .3532399
              |
        party |
           2  |  -.1919573    .086694    -2.21   0.027    -.3618744   -.0220402
           3  |   .7756426   .0927817     8.36   0.000     .5937937    .9574915
              |
outcome#party |
         2 2  |    .640166   .1226601     5.22   0.000     .3997565    .8805754
         2 3  |  -.7512935   .1307307    -5.75   0.000    -1.007521    -.495066
              |
        _cons |  -.3085721   .0618708    -4.99   0.000    -.4298366   -.1873076
-------------------------------------------------------------------------------

.         margins i.outcome, at(party=3) vsquish post

Adjusted predictions                            Number of obs     =      2,466
Model VCE    : Robust

Expression   : Pr(trust_b), predict()
at           : party           =           3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     outcome |
          1  |   .6797753   .0247328    27.48   0.000     .6312999    .7282507
          2  |   .4581006   .0263382    17.39   0.000     .4064786    .5097225
------------------------------------------------------------------------------

.         estimates store rep4                    

. 
. // generate partisan figure for procedural legitimacy variable
. coefplot (dem3, color(538b) ciopts(lcolor(black) recast(rcap))) ///
>          (rep3, color(538r) ciopts(lcolor(black) recast(rcap))) ///
>                  (ind3, color(538g) ciopts(lcolor(black) recast(rcap))), ///
>                   bylabel("(a) UNHCR is Fair") || ///
>                  (dem4, color(538b) ciopts(lcolor(black) recast(rcap))) ///
>          (rep4, color(538r) ciopts(lcolor(black) recast(rcap))) ///
>                  (ind4, color(538g) ciopts(lcolor(black) recast(rcap))), ///
>                   bylabel("(b) UNHCR can be Trusted") vertical ///
>                  ciopts(recast(rcap) lcol(black)) citop scale(1.25) ///
>                  ytitle("Mean Percent") format(%9.2g) rescale(100) ///
>                  xlabel(1 `""Non-critical" "Report of the U.S.""' ///
>                             2 `""Critical" "Report of the U.S.""', labcol(black)) ///
>                  legend(span rows(1) label(1 "Democrat") label(3 "Republican") ///
>              label(5 "Independent") size(medsmall)) ///
>                  recast(bar) barwidth(.2) fysize(140) fxsize(100) ///
>                  addplot(scatter @b @at, ms(i) mlabel(@b) mlabpos(2) mlabcolor(black))

. graph save "outcome1.gph", replace               
(note: file outcome1.gph not found)
(file outcome1.gph saved)

. 
. // generate partisan figure for substantive legitimacy variable          
. coefplot (dem1, color(538b) ciopts(lcolor(black) recast(rcap))) ///
>          (rep1, color(538r) ciopts(lcolor(black) recast(rcap))) ///
>                  (ind1, color(538g) ciopts(lcolor(black) recast(rcap))), ///
>                   bylabel("(c) Outcome Good for Americans") || ///
>                  (dem2, color(538b) ciopts(lcolor(black) recast(rcap))) ///
>          (rep2, color(538r) ciopts(lcolor(black) recast(rcap))) ///
>                  (ind2, color(538g) ciopts(lcolor(black) recast(rcap))), /// 
>                   bylabel("(d) Outcome Good for Refugees") vertical ///
>                  ciopts(recast(rcap) lcol(black)) citop scale(1.25) ///
>                  ytitle("Mean Percent") format(%9.2g) rescale(100) ///
>                  xlabel(1 `""Non-critical" "Report of the U.S.""' ///
>                             2 `""Critical" "Report of the U.S.""', labcol(black)) ///
>                  fysize(120) fxsize(100) ///
>                  byopts(legend(off)) recast(bar) barwidth(.2) ///
>                  addplot(scatter @b @at, ms(i) mlabel(@b) mlabpos(2) mlabcolor(black))

. graph save "outcome2.gph", replace              
(note: file outcome2.gph not found)
(file outcome2.gph saved)

. 
. // combine saved .gph files into a single figure         
. graph combine "outcome1.gph" "outcome2.gph", xcommon rows(2) 
(note:  clockdir by_legend_position not found in scheme, default attributes used)
(note:  clockdir by_legend_position not found in scheme, default attributes used)
(note:  clockdir by_legend_position not found in scheme, default attributes used)

.                                  
. ********************************************************************************
. *** Figure 7 *******************************************************************
. ********************************************************************************
. 
. // analysis for partisan effects on american outcome
. reg american i.panel##i.party, robust

Linear regression                               Number of obs     =      2,466
                                                F(11, 2454)       =       2.26
                                                Prob > F          =     0.0097
                                                R-squared         =     0.0097
                                                Root MSE          =     1.4031

------------------------------------------------------------------------------
             |               Robust
    american |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |     .15671   .1353329     1.16   0.247    -.1086685    .4220884
          3  |   .0467721   .1333846     0.35   0.726    -.2147859      .30833
          4  |   .3636364   .1357043     2.68   0.007     .0975295    .6297432
             |
       party |
          2  |   .1174242   .1388122     0.85   0.398    -.1547769    .3896254
          3  |   .4128153   .1443886     2.86   0.004     .1296792    .6959514
             |
 panel#party |
        2 2  |  -.0527184   .1922263    -0.27   0.784    -.4296609    .3242242
        2 3  |  -.1907374   .2038517    -0.94   0.350    -.5904766    .2090018
        3 2  |    .049339   .1900563     0.26   0.795    -.3233482    .4220263
        3 3  |  -.1677126   .2009561    -0.83   0.404    -.5617737    .2263486
        4 2  |  -.0818567     .18862    -0.43   0.664    -.4517275    .2880141
        4 3  |  -.3730105   .2033671    -1.83   0.067    -.7717994    .0257785
             |
       _cons |   3.257576   .0981543    33.19   0.000     3.065102     3.45005
------------------------------------------------------------------------------

.         margins r.panel, at(party=1) vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()
at           : party           =           1

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        1.34     0.2470
   (3 vs 1)  |          1        0.12     0.7259
   (4 vs 1)  |          1        7.18     0.0074
      Joint  |          3        2.94     0.0319
             |
 Denominator |       2454
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |     .15671   .1353329     -.1086685    .4220884
   (3 vs 1)  |   .0467721   .1333846     -.2147859      .30833
   (4 vs 1)  |   .3636364   .1357043      .0975295    .6297432
--------------------------------------------------------------

.         estimates store ind1

. reg american i.panel##i.party, robust

Linear regression                               Number of obs     =      2,466
                                                F(11, 2454)       =       2.26
                                                Prob > F          =     0.0097
                                                R-squared         =     0.0097
                                                Root MSE          =     1.4031

------------------------------------------------------------------------------
             |               Robust
    american |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |     .15671   .1353329     1.16   0.247    -.1086685    .4220884
          3  |   .0467721   .1333846     0.35   0.726    -.2147859      .30833
          4  |   .3636364   .1357043     2.68   0.007     .0975295    .6297432
             |
       party |
          2  |   .1174242   .1388122     0.85   0.398    -.1547769    .3896254
          3  |   .4128153   .1443886     2.86   0.004     .1296792    .6959514
             |
 panel#party |
        2 2  |  -.0527184   .1922263    -0.27   0.784    -.4296609    .3242242
        2 3  |  -.1907374   .2038517    -0.94   0.350    -.5904766    .2090018
        3 2  |    .049339   .1900563     0.26   0.795    -.3233482    .4220263
        3 3  |  -.1677126   .2009561    -0.83   0.404    -.5617737    .2263486
        4 2  |  -.0818567     .18862    -0.43   0.664    -.4517275    .2880141
        4 3  |  -.3730105   .2033671    -1.83   0.067    -.7717994    .0257785
             |
       _cons |   3.257576   .0981543    33.19   0.000     3.065102     3.45005
------------------------------------------------------------------------------

.         margins r.panel, at(party=2) vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()
at           : party           =           2

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        0.58     0.4463
   (3 vs 1)  |          1        0.50     0.4778
   (4 vs 1)  |          1        4.63     0.0316
      Joint  |          3        1.67     0.1705
             |
 Denominator |       2454
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .1039916   .1365136     -.1637022    .3716854
   (3 vs 1)  |   .0961111   .1353882     -.1693757     .361598
   (4 vs 1)  |   .2817797   .1310032      .0248914    .5386679
--------------------------------------------------------------

.         estimates store dem1

. reg american i.panel##i.party, robust

Linear regression                               Number of obs     =      2,466
                                                F(11, 2454)       =       2.26
                                                Prob > F          =     0.0097
                                                R-squared         =     0.0097
                                                Root MSE          =     1.4031

------------------------------------------------------------------------------
             |               Robust
    american |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |     .15671   .1353329     1.16   0.247    -.1086685    .4220884
          3  |   .0467721   .1333846     0.35   0.726    -.2147859      .30833
          4  |   .3636364   .1357043     2.68   0.007     .0975295    .6297432
             |
       party |
          2  |   .1174242   .1388122     0.85   0.398    -.1547769    .3896254
          3  |   .4128153   .1443886     2.86   0.004     .1296792    .6959514
             |
 panel#party |
        2 2  |  -.0527184   .1922263    -0.27   0.784    -.4296609    .3242242
        2 3  |  -.1907374   .2038517    -0.94   0.350    -.5904766    .2090018
        3 2  |    .049339   .1900563     0.26   0.795    -.3233482    .4220263
        3 3  |  -.1677126   .2009561    -0.83   0.404    -.5617737    .2263486
        4 2  |  -.0818567     .18862    -0.43   0.664    -.4517275    .2880141
        4 3  |  -.3730105   .2033671    -1.83   0.067    -.7717994    .0257785
             |
       _cons |   3.257576   .0981543    33.19   0.000     3.065102     3.45005
------------------------------------------------------------------------------

.         margins r.panel, at(party=3) vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()
at           : party           =           3

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        0.05     0.8234
   (3 vs 1)  |          1        0.65     0.4211
   (4 vs 1)  |          1        0.00     0.9507
      Joint  |          3        0.27     0.8493
             |
 Denominator |       2454
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |  -.0340274   .1524485     -.3329684    .2649135
   (3 vs 1)  |  -.1209405   .1503061     -.4156804    .1737994
   (4 vs 1)  |  -.0093741   .1514679     -.3063922     .287644
--------------------------------------------------------------

.         estimates store rep1    

. // analysis for partisan effects on refugee outcome
. reg refugee i.panel##i.party, robust

Linear regression                               Number of obs     =      2,466
                                                F(11, 2454)       =       3.35
                                                Prob > F          =     0.0001
                                                R-squared         =     0.0142
                                                Root MSE          =     1.3858

------------------------------------------------------------------------------
             |               Robust
     refugee |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .0972583   .1349735     0.72   0.471    -.1674155    .3619321
          3  |   .0160299   .1352199     0.12   0.906    -.2491271    .2811868
          4  |   .2878788   .1350421     2.13   0.033     .0230705    .5526871
             |
       party |
          2  |   .0020202   .1428537     0.01   0.989    -.2781061    .2821465
          3  |   .5009029   .1381028     3.63   0.000     .2300929    .7717129
             |
 panel#party |
        2 2  |   .0775316   .1944163     0.40   0.690    -.3037054    .4587686
        2 3  |  -.1745501   .1939792    -0.90   0.368    -.5549299    .2058297
        3 2  |   .0661924   .1953861     0.34   0.735    -.3169464    .4493311
        3 3  |  -.2819455   .1935415    -1.46   0.145    -.6614671     .097576
        4 2  |  -.0073703   .1926281    -0.04   0.969    -.3851007    .3703601
        4 3  |  -.3522982   .1921657    -1.83   0.067    -.7291219    .0245255
             |
       _cons |    3.29798   .0991359    33.27   0.000     3.103581    3.492378
------------------------------------------------------------------------------

.         margins r.panel, at(party=1) vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()
at           : party           =           1

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        0.52     0.4712
   (3 vs 1)  |          1        0.01     0.9056
   (4 vs 1)  |          1        4.54     0.0331
      Joint  |          3        2.01     0.1099
             |
 Denominator |       2454
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .0972583   .1349735     -.1674155    .3619321
   (3 vs 1)  |   .0160299   .1352199     -.2491271    .2811868
   (4 vs 1)  |   .2878788   .1350421      .0230705    .5526871
--------------------------------------------------------------

.         estimates store ind2

. reg refugee i.panel##i.party, robust

Linear regression                               Number of obs     =      2,466
                                                F(11, 2454)       =       3.35
                                                Prob > F          =     0.0001
                                                R-squared         =     0.0142
                                                Root MSE          =     1.3858

------------------------------------------------------------------------------
             |               Robust
     refugee |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .0972583   .1349735     0.72   0.471    -.1674155    .3619321
          3  |   .0160299   .1352199     0.12   0.906    -.2491271    .2811868
          4  |   .2878788   .1350421     2.13   0.033     .0230705    .5526871
             |
       party |
          2  |   .0020202   .1428537     0.01   0.989    -.2781061    .2821465
          3  |   .5009029   .1381028     3.63   0.000     .2300929    .7717129
             |
 panel#party |
        2 2  |   .0775316   .1944163     0.40   0.690    -.3037054    .4587686
        2 3  |  -.1745501   .1939792    -0.90   0.368    -.5549299    .2058297
        3 2  |   .0661924   .1953861     0.34   0.735    -.3169464    .4493311
        3 3  |  -.2819455   .1935415    -1.46   0.145    -.6614671     .097576
        4 2  |  -.0073703   .1926281    -0.04   0.969    -.3851007    .3703601
        4 3  |  -.3522982   .1921657    -1.83   0.067    -.7291219    .0245255
             |
       _cons |    3.29798   .0991359    33.27   0.000     3.103581    3.492378
------------------------------------------------------------------------------

.         margins r.panel, at(party=2) vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()
at           : party           =           2

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        1.56     0.2117
   (3 vs 1)  |          1        0.34     0.5600
   (4 vs 1)  |          1        4.17     0.0413
      Joint  |          3        1.57     0.1936
             |
 Denominator |       2454
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .1747899    .139928     -.0995993    .4491791
   (3 vs 1)  |   .0822222   .1410366     -.1943408    .3587852
   (4 vs 1)  |   .2805085   .1373652      .0111447    .5498723
--------------------------------------------------------------

.         estimates store dem2

. reg refugee i.panel##i.party, robust

Linear regression                               Number of obs     =      2,466
                                                F(11, 2454)       =       3.35
                                                Prob > F          =     0.0001
                                                R-squared         =     0.0142
                                                Root MSE          =     1.3858

------------------------------------------------------------------------------
             |               Robust
     refugee |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .0972583   .1349735     0.72   0.471    -.1674155    .3619321
          3  |   .0160299   .1352199     0.12   0.906    -.2491271    .2811868
          4  |   .2878788   .1350421     2.13   0.033     .0230705    .5526871
             |
       party |
          2  |   .0020202   .1428537     0.01   0.989    -.2781061    .2821465
          3  |   .5009029   .1381028     3.63   0.000     .2300929    .7717129
             |
 panel#party |
        2 2  |   .0775316   .1944163     0.40   0.690    -.3037054    .4587686
        2 3  |  -.1745501   .1939792    -0.90   0.368    -.5549299    .2058297
        3 2  |   .0661924   .1953861     0.34   0.735    -.3169464    .4493311
        3 3  |  -.2819455   .1935415    -1.46   0.145    -.6614671     .097576
        4 2  |  -.0073703   .1926281    -0.04   0.969    -.3851007    .3703601
        4 3  |  -.3522982   .1921657    -1.83   0.067    -.7291219    .0245255
             |
       _cons |    3.29798   .0991359    33.27   0.000     3.103581    3.492378
------------------------------------------------------------------------------

.         margins r.panel, at(party=3) vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()
at           : party           =           3

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        0.31     0.5791
   (3 vs 1)  |          1        3.69     0.0549
   (4 vs 1)  |          1        0.22     0.6375
      Joint  |          3        1.35     0.2576
             |
 Denominator |       2454
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |  -.0772918     .13932     -.3504887    .1959052
   (3 vs 1)  |  -.2659156   .1384698     -.5374453    .0056141
   (4 vs 1)  |  -.0644194   .1367161     -.3325102    .2036714
--------------------------------------------------------------

.         estimates store rep2    

. // analysis for partisan effects on fairness outcome
. reg fair i.panel##i.party, robust

Linear regression                               Number of obs     =      2,466
                                                F(11, 2454)       =       9.01
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0389
                                                Root MSE          =     1.2326

------------------------------------------------------------------------------
             |               Robust
        fair |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .2458874    .115351     2.13   0.033     .0196921    .4720828
          3  |   .1991656   .1181179     1.69   0.092    -.0324556    .4307867
          4  |    .520202   .1183174     4.40   0.000     .2881897    .7522143
             |
       party |
          2  |   .0530303   .1204715     0.44   0.660     -.183206    .2892666
          3  |   .5665312   .1238725     4.57   0.000     .3236259    .8094366
             |
 panel#party |
        2 2  |   -.229781   .1664726    -1.38   0.168    -.5562224    .0966604
        2 3  |  -.1306005   .1756299    -0.74   0.457    -.4749987    .2137977
        3 2  |  -.0224989   .1708267    -0.13   0.895    -.3574783    .3124805
        3 3  |  -.1897727    .173442    -1.09   0.274    -.5298805     .150335
        4 2  |   .0716059   .1654391     0.43   0.665    -.2528087    .3960205
        4 3  |  -.4644939    .176082    -2.64   0.008    -.8097785   -.1192093
             |
       _cons |   2.863636   .0837006    34.21   0.000     2.699505    3.027767
------------------------------------------------------------------------------

.         margins r.panel, at(party=1) vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()
at           : party           =           1

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        4.54     0.0331
   (3 vs 1)  |          1        2.84     0.0919
   (4 vs 1)  |          1       19.33     0.0000
      Joint  |          3        6.56     0.0002
             |
 Denominator |       2454
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .2458874    .115351      .0196921    .4720828
   (3 vs 1)  |   .1991656   .1181179     -.0324556    .4307867
   (4 vs 1)  |    .520202   .1183174      .2881897    .7522143
--------------------------------------------------------------

.         estimates store ind3

. reg fair i.panel##i.party, robust

Linear regression                               Number of obs     =      2,466
                                                F(11, 2454)       =       9.01
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0389
                                                Root MSE          =     1.2326

------------------------------------------------------------------------------
             |               Robust
        fair |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .2458874    .115351     2.13   0.033     .0196921    .4720828
          3  |   .1991656   .1181179     1.69   0.092    -.0324556    .4307867
          4  |    .520202   .1183174     4.40   0.000     .2881897    .7522143
             |
       party |
          2  |   .0530303   .1204715     0.44   0.660     -.183206    .2892666
          3  |   .5665312   .1238725     4.57   0.000     .3236259    .8094366
             |
 panel#party |
        2 2  |   -.229781   .1664726    -1.38   0.168    -.5562224    .0966604
        2 3  |  -.1306005   .1756299    -0.74   0.457    -.4749987    .2137977
        3 2  |  -.0224989   .1708267    -0.13   0.895    -.3574783    .3124805
        3 3  |  -.1897727    .173442    -1.09   0.274    -.5298805     .150335
        4 2  |   .0716059   .1654391     0.43   0.665    -.2528087    .3960205
        4 3  |  -.4644939    .176082    -2.64   0.008    -.8097785   -.1192093
             |
       _cons |   2.863636   .0837006    34.21   0.000     2.699505    3.027767
------------------------------------------------------------------------------

.         margins r.panel, at(party=2) vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()
at           : party           =           2

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        0.02     0.8933
   (3 vs 1)  |          1        2.05     0.1524
   (4 vs 1)  |          1       26.19     0.0000
      Joint  |          3       12.10     0.0000
             |
 Denominator |       2454
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .0161064   .1200304     -.2192649    .2514777
   (3 vs 1)  |   .1766667   .1234096      -.065331    .4186643
   (4 vs 1)  |   .5918079   .1156334      .3650588     .818557
--------------------------------------------------------------

.         estimates store dem3

. reg fair i.panel##i.party, robust

Linear regression                               Number of obs     =      2,466
                                                F(11, 2454)       =       9.01
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0389
                                                Root MSE          =     1.2326

------------------------------------------------------------------------------
             |               Robust
        fair |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .2458874    .115351     2.13   0.033     .0196921    .4720828
          3  |   .1991656   .1181179     1.69   0.092    -.0324556    .4307867
          4  |    .520202   .1183174     4.40   0.000     .2881897    .7522143
             |
       party |
          2  |   .0530303   .1204715     0.44   0.660     -.183206    .2892666
          3  |   .5665312   .1238725     4.57   0.000     .3236259    .8094366
             |
 panel#party |
        2 2  |   -.229781   .1664726    -1.38   0.168    -.5562224    .0966604
        2 3  |  -.1306005   .1756299    -0.74   0.457    -.4749987    .2137977
        3 2  |  -.0224989   .1708267    -0.13   0.895    -.3574783    .3124805
        3 3  |  -.1897727    .173442    -1.09   0.274    -.5298805     .150335
        4 2  |   .0716059   .1654391     0.43   0.665    -.2528087    .3960205
        4 3  |  -.4644939    .176082    -2.64   0.008    -.8097785   -.1192093
             |
       _cons |   2.863636   .0837006    34.21   0.000     2.699505    3.027767
------------------------------------------------------------------------------

.         margins r.panel, at(party=3) vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()
at           : party           =           3

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        0.76     0.3841
   (3 vs 1)  |          1        0.01     0.9411
   (4 vs 1)  |          1        0.18     0.6693
      Joint  |          3        0.32     0.8126
             |
 Denominator |       2454
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .1152869   .1324388     -.1444164    .3749903
   (3 vs 1)  |   .0093928    .127005     -.2396552    .2584409
   (4 vs 1)  |   .0557081   .1304065       -.20001    .3114262
--------------------------------------------------------------

.         estimates store rep3            

. // analysis for partisan effects on trust outcome       
. reg trust i.panel##i.party, robust

Linear regression                               Number of obs     =      2,466
                                                F(11, 2454)       =       7.26
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0323
                                                Root MSE          =     1.2736

------------------------------------------------------------------------------
             |               Robust
       trust |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .2210678   .1237272     1.79   0.074    -.0215526    .4636883
          3  |   .2334212   .1265037     1.85   0.065    -.0146439    .4814862
          4  |   .4343434   .1264932     3.43   0.001      .186299    .6823879
             |
       party |
          2  |   .0061869   .1295999     0.05   0.962    -.2479497    .2603234
          3  |   .5884262   .1349373     4.36   0.000     .3238233     .853029
             |
 panel#party |
        2 2  |   .0938131   .1757124     0.53   0.593    -.2507467     .438373
        2 3  |  -.1093677   .1850041    -0.59   0.554     -.472148    .2534125
        3 2  |   .0001899   .1782048     0.00   0.999    -.3492575    .3496374
        3 3  |  -.3139663   .1857734    -1.69   0.091    -.6782552    .0503226
        4 2  |   .1583826   .1733697     0.91   0.361    -.1815834    .4983485
        4 3  |  -.3433483   .1858435    -1.85   0.065    -.7077745     .021078
             |
       _cons |   2.964646   .0920758    32.20   0.000     2.784092    3.145201
------------------------------------------------------------------------------

.         margins r.panel, at(party=1) vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()
at           : party           =           1

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        3.19     0.0741
   (3 vs 1)  |          1        3.40     0.0651
   (4 vs 1)  |          1       11.79     0.0006
      Joint  |          3        3.93     0.0082
             |
 Denominator |       2454
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .2210678   .1237272     -.0215526    .4636883
   (3 vs 1)  |   .2334212   .1265037     -.0146439    .4814862
   (4 vs 1)  |   .4343434   .1264932       .186299    .6823879
--------------------------------------------------------------

.         estimates store ind4

. reg trust i.panel##i.party, robust

Linear regression                               Number of obs     =      2,466
                                                F(11, 2454)       =       7.26
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0323
                                                Root MSE          =     1.2736

------------------------------------------------------------------------------
             |               Robust
       trust |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .2210678   .1237272     1.79   0.074    -.0215526    .4636883
          3  |   .2334212   .1265037     1.85   0.065    -.0146439    .4814862
          4  |   .4343434   .1264932     3.43   0.001      .186299    .6823879
             |
       party |
          2  |   .0061869   .1295999     0.05   0.962    -.2479497    .2603234
          3  |   .5884262   .1349373     4.36   0.000     .3238233     .853029
             |
 panel#party |
        2 2  |   .0938131   .1757124     0.53   0.593    -.2507467     .438373
        2 3  |  -.1093677   .1850041    -0.59   0.554     -.472148    .2534125
        3 2  |   .0001899   .1782048     0.00   0.999    -.3492575    .3496374
        3 3  |  -.3139663   .1857734    -1.69   0.091    -.6782552    .0503226
        4 2  |   .1583826   .1733697     0.91   0.361    -.1815834    .4983485
        4 3  |  -.3433483   .1858435    -1.85   0.065    -.7077745     .021078
             |
       _cons |   2.964646   .0920758    32.20   0.000     2.784092    3.145201
------------------------------------------------------------------------------

.         margins r.panel, at(party=2) vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()
at           : party           =           2

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        6.37     0.0117
   (3 vs 1)  |          1        3.46     0.0628
   (4 vs 1)  |          1       24.99     0.0000
      Joint  |          3        8.75     0.0000
             |
 Denominator |       2454
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |    .314881   .1247655      .0702244    .5595375
   (3 vs 1)  |   .2336111   .1255141     -.0125134    .4797356
   (4 vs 1)  |    .592726   .1185602      .3602377    .8252143
--------------------------------------------------------------

.         estimates store dem4

. reg trust i.panel##i.party, robust

Linear regression                               Number of obs     =      2,466
                                                F(11, 2454)       =       7.26
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0323
                                                Root MSE          =     1.2736

------------------------------------------------------------------------------
             |               Robust
       trust |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .2210678   .1237272     1.79   0.074    -.0215526    .4636883
          3  |   .2334212   .1265037     1.85   0.065    -.0146439    .4814862
          4  |   .4343434   .1264932     3.43   0.001      .186299    .6823879
             |
       party |
          2  |   .0061869   .1295999     0.05   0.962    -.2479497    .2603234
          3  |   .5884262   .1349373     4.36   0.000     .3238233     .853029
             |
 panel#party |
        2 2  |   .0938131   .1757124     0.53   0.593    -.2507467     .438373
        2 3  |  -.1093677   .1850041    -0.59   0.554     -.472148    .2534125
        3 2  |   .0001899   .1782048     0.00   0.999    -.3492575    .3496374
        3 3  |  -.3139663   .1857734    -1.69   0.091    -.6782552    .0503226
        4 2  |   .1583826   .1733697     0.91   0.361    -.1815834    .4983485
        4 3  |  -.3433483   .1858435    -1.85   0.065    -.7077745     .021078
             |
       _cons |   2.964646   .0920758    32.20   0.000     2.784092    3.145201
------------------------------------------------------------------------------

.         margins r.panel, at(party=3) vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()
at           : party           =           3

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        0.66     0.4168
   (3 vs 1)  |          1        0.35     0.5539
   (4 vs 1)  |          1        0.45     0.5040
      Joint  |          3        0.88     0.4518
             |
 Denominator |       2454
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .1117001   .1375431     -.1580124    .3814126
   (3 vs 1)  |  -.0805452   .1360462     -.3473225    .1862322
   (4 vs 1)  |   .0909952   .1361516     -.1759888    .3579791
--------------------------------------------------------------

.         estimates store rep4    

. 
. // generate figure for diversity and legitimacy outcome conditional by partyid
. coefplot (dem3, color(538b) ciopts(lcolor(538b) recast(rcap))) ///
>          (rep3, color(538r) ciopts(lcolor(538r) recast(rcap))) ///
>                  (ind3, color(538g) ciopts(lcolor(538g) recast(rcap))), ///
>                   bylabel("(a) UNHCR Panel Is Fair") || ///
>                  (dem4, color(538b) ciopts(lcolor(538b) recast(rcap))) ///
>          (rep4, color(538r) ciopts(lcolor(538r) recast(rcap))) ///
>                  (ind4, color(538g) ciopts(lcolor(538g) recast(rcap))), ///
>                   bylabel("(b) UNHCR Panel Can Be Trusted") || ///
>                  (dem1, color(538b) ciopts(lcolor(538b) recast(rcap))) ///
>          (rep1, color(538r) ciopts(lcolor(538r) recast(rcap))) ///
>                  (ind1, color(538g) ciopts(lcolor(538g) recast(rcap))), ///
>                   bylabel("(c) Outcome Good for American") || ///
>                  (dem2, color(538b) ciopts(lcolor(538b) recast(rcap))) ///
>          (rep2, color(538r) ciopts(lcolor(538r) recast(rcap))) ///
>                  (ind2, color(538g) ciopts(lcolor(538g) recast(rcap))), ///
>                   bylabel("(d) Outcome Good for Refugee") vertical ///
>                  yline(0) ytitle("Marginal Difference of White-Male vs. Diverse Panels") ///
>                  xtitle("Gender and Racial Distribution of Panel (All-White, All-Male as Reference 
> Group)") ///
>                  xlabel(1 `""Mixed-Race" "All-Male""' 2 `""All-White" "Mixed-Gender""' ///
>                                 3 `""Mixed-Race" "Mixed-Gender""', labcol(black)) ///
>                  legend(order(2 "Democrats" 4 "Republicans" 6 "Independents") rows(1))  

.                                 
.                                 
. ********************************************************************************
. *** Figure 8 *******************************************************************
. ********************************************************************************
. 
. // Input main Sample analyses
. use "CH_main.dta", clear        

. 
. // analyses of main results
. reg american i.panel, robust

Linear regression                               Number of obs     =      2,557
                                                F(3, 2553)        =       4.10
                                                Prob > F          =     0.0065
                                                R-squared         =     0.0046
                                                Root MSE          =     1.4017

------------------------------------------------------------------------------
             |               Robust
    american |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .0817817   .0800913     1.02   0.307    -.0752689    .2388323
          3  |   .0133721   .0791005     0.17   0.866    -.1417355    .1684797
          4  |   .2407824   .0786437     3.06   0.002     .0865704    .3949944
             |
       _cons |   3.417445   .0573209    59.62   0.000     3.305045    3.529846
------------------------------------------------------------------------------

.         margins r.panel, vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        1.04     0.3073
   (3 vs 1)  |          1        0.03     0.8658
   (4 vs 1)  |          1        9.37     0.0022
      Joint  |          3        4.10     0.0065
             |
 Denominator |       2553
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .0817817   .0800913     -.0752689    .2388323
   (3 vs 1)  |   .0133721   .0791005     -.1417355    .1684797
   (4 vs 1)  |   .2407824   .0786437      .0865704    .3949944
--------------------------------------------------------------

.         estimates store american1

. reg refugee i.panel, robust

Linear regression                               Number of obs     =      2,557
                                                F(3, 2553)        =       3.32
                                                Prob > F          =     0.0190
                                                R-squared         =     0.0037
                                                Root MSE          =     1.3875

------------------------------------------------------------------------------
             |               Robust
     refugee |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |     .07851   .0792808     0.99   0.322    -.0769513    .2339712
          3  |  -.0382942   .0789904    -0.48   0.628    -.1931861    .1165976
          4  |     .18419   .0782573     2.35   0.019     .0307357    .3376444
             |
       _cons |    3.44081   .0574929    59.85   0.000     3.328072    3.553547
------------------------------------------------------------------------------

.         margins r.panel, vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        0.98     0.3221
   (3 vs 1)  |          1        0.24     0.6279
   (4 vs 1)  |          1        5.54     0.0187
      Joint  |          3        3.32     0.0190
             |
 Denominator |       2553
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |     .07851   .0792808     -.0769513    .2339712
   (3 vs 1)  |  -.0382942   .0789904     -.1931861    .1165976
   (4 vs 1)  |     .18419   .0782573      .0307357    .3376444
--------------------------------------------------------------

.         estimates store refugee1

. reg fair i.panel, robust

Linear regression                               Number of obs     =      2,556
                                                F(3, 2552)        =      13.11
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0146
                                                Root MSE          =     1.2431

------------------------------------------------------------------------------
             |               Robust
        fair |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .1325841   .0706638     1.88   0.061      -.00598    .2711483
          3  |   .1374887   .0703826     1.95   0.051    -.0005242    .2755016
          4  |   .4168293   .0691973     6.02   0.000     .2811407     .552518
             |
       _cons |   3.043614   .0503386    60.46   0.000     2.944905    3.142322
------------------------------------------------------------------------------

.         margins r.panel, vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        3.52     0.0607
   (3 vs 1)  |          1        3.82     0.0509
   (4 vs 1)  |          1       36.29     0.0000
      Joint  |          3       13.11     0.0000
             |
 Denominator |       2552
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .1325841   .0706638       -.00598    .2711483
   (3 vs 1)  |   .1374887   .0703826     -.0005242    .2755016
   (4 vs 1)  |   .4168293   .0691973      .2811407     .552518
--------------------------------------------------------------

.         estimates store fair1

. reg trust i.panel, robust

Linear regression                               Number of obs     =      2,554
                                                F(3, 2550)        =      11.01
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0127
                                                Root MSE          =     1.2836

------------------------------------------------------------------------------
             |               Robust
       trust |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .2174272   .0740354     2.94   0.003     .0722514    .3626029
          3  |   .1448009   .0737736     1.96   0.050     .0001387    .2894632
          4  |    .404898   .0724889     5.59   0.000     .2627549    .5470411
             |
       _cons |   3.135514   .0541623    57.89   0.000     3.029308    3.241721
------------------------------------------------------------------------------

.         margins r.panel, vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        8.62     0.0033
   (3 vs 1)  |          1        3.85     0.0498
   (4 vs 1)  |          1       31.20     0.0000
      Joint  |          3       11.01     0.0000
             |
 Denominator |       2550
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .2174272   .0740354      .0722514    .3626029
   (3 vs 1)  |   .1448009   .0737736      .0001387    .2894632
   (4 vs 1)  |    .404898   .0724889      .2627549    .5470411
--------------------------------------------------------------

.         estimates store trust1  

.         
.         
. // input follow-up survey data
. use "CH_followup.dta", clear 

. 
. // analyses of follow-up sample w/o country labels analyses
. reg american i.panel if country==1, robust

Linear regression                               Number of obs     =      1,428
                                                F(3, 1424)        =       0.70
                                                Prob > F          =     0.5530
                                                R-squared         =     0.0014
                                                Root MSE          =     1.2195

------------------------------------------------------------------------------
             |               Robust
    american |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .0364141   .0936266     0.39   0.697    -.1472469     .220075
          3  |   .0593473   .0921759     0.64   0.520    -.1214678    .2401624
          4  |    .126078   .0904183     1.39   0.163    -.0512895    .3034454
             |
       _cons |   3.484419   .0661758    52.65   0.000     3.354607    3.614232
------------------------------------------------------------------------------

.         margins r.panel, vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        0.15     0.6974
   (3 vs 1)  |          1        0.41     0.5198
   (4 vs 1)  |          1        1.94     0.1634
      Joint  |          3        0.70     0.5530
             |
 Denominator |       1424
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .0364141   .0936266     -.1472469     .220075
   (3 vs 1)  |   .0593473   .0921759     -.1214678    .2401624
   (4 vs 1)  |    .126078   .0904183     -.0512895    .3034454
--------------------------------------------------------------

.         estimates store american2

. reg refugee i.panel if country==1, robust

Linear regression                               Number of obs     =      1,428
                                                F(3, 1424)        =       1.92
                                                Prob > F          =     0.1250
                                                R-squared         =     0.0041
                                                Root MSE          =     1.2433

------------------------------------------------------------------------------
             |               Robust
     refugee |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .1369638   .0966002     1.42   0.156    -.0525302    .3264578
          3  |   .1206934   .0961304     1.26   0.209    -.0678789    .3092657
          4  |   .2232404   .0933613     2.39   0.017        .0401    .4063808
             |
       _cons |   3.354108   .0705647    47.53   0.000     3.215686     3.49253
------------------------------------------------------------------------------

.         margins r.panel, vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        2.01     0.1565
   (3 vs 1)  |          1        1.58     0.2095
   (4 vs 1)  |          1        5.72     0.0169
      Joint  |          3        1.92     0.1250
             |
 Denominator |       1424
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .1369638   .0966002     -.0525302    .3264578
   (3 vs 1)  |   .1206934   .0961304     -.0678789    .3092657
   (4 vs 1)  |   .2232404   .0933613         .0401    .4063808
--------------------------------------------------------------

.         estimates store refugee2

. reg fair i.panel if country==1, robust

Linear regression                               Number of obs     =      1,428
                                                F(3, 1424)        =      11.85
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0225
                                                Root MSE          =     1.1045

------------------------------------------------------------------------------
             |               Robust
        fair |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .2042358   .0870784     2.35   0.019     .0334201    .3750515
          3  |   .2187615   .0863581     2.53   0.011     .0493588    .3881642
          4  |   .4711392   .0808547     5.83   0.000     .3125321    .6297464
             |
       _cons |   3.271955   .0627892    52.11   0.000     3.148785    3.395124
------------------------------------------------------------------------------

.         margins r.panel, vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        5.50     0.0191
   (3 vs 1)  |          1        6.42     0.0114
   (4 vs 1)  |          1       33.95     0.0000
      Joint  |          3       11.85     0.0000
             |
 Denominator |       1424
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .2042358   .0870784      .0334201    .3750515
   (3 vs 1)  |   .2187615   .0863581      .0493588    .3881642
   (4 vs 1)  |   .4711392   .0808547      .3125321    .6297464
--------------------------------------------------------------

.         estimates store fair2

. reg trust i.panel if country==1, robust

Linear regression                               Number of obs     =      1,428
                                                F(3, 1424)        =       7.49
                                                Prob > F          =     0.0001
                                                R-squared         =     0.0148
                                                Root MSE          =     1.1591

------------------------------------------------------------------------------
             |               Robust
       trust |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .1883178   .0905121     2.08   0.038     .0107664    .3658692
          3  |   .1203628   .0880783     1.37   0.172    -.0524144      .29314
          4  |   .3902384    .085675     4.55   0.000     .2221756    .5583011
             |
       _cons |   3.192635   .0639957    49.89   0.000     3.067099    3.318171
------------------------------------------------------------------------------

.         margins r.panel, vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        4.33     0.0377
   (3 vs 1)  |          1        1.87     0.1720
   (4 vs 1)  |          1       20.75     0.0000
      Joint  |          3        7.49     0.0001
             |
 Denominator |       1424
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .1883178   .0905121      .0107664    .3658692
   (3 vs 1)  |   .1203628   .0880783     -.0524144      .29314
   (4 vs 1)  |   .3902384    .085675      .2221756    .5583011
--------------------------------------------------------------

.         estimates store trust2          

. // analyses of follow-up sample w country labels analyses
. reg american i.panel if country==2, robust

Linear regression                               Number of obs     =      1,478
                                                F(3, 1474)        =       0.58
                                                Prob > F          =     0.6275
                                                R-squared         =     0.0012
                                                Root MSE          =     1.2707

------------------------------------------------------------------------------
             |               Robust
    american |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |  -.0080982   .0927522    -0.09   0.930    -.1900385    .1738421
          3  |    .056044   .0921142     0.61   0.543    -.1246449    .2367329
          4  |   .1000573   .0932533     1.07   0.283    -.0828659    .2829805
             |
       _cons |   3.441253   .0643842    53.45   0.000     3.314959    3.567548
------------------------------------------------------------------------------

.         margins r.panel, vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        0.01     0.9304
   (3 vs 1)  |          1        0.37     0.5430
   (4 vs 1)  |          1        1.15     0.2835
      Joint  |          3        0.58     0.6275
             |
 Denominator |       1474
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |  -.0080982   .0927522     -.1900385    .1738421
   (3 vs 1)  |    .056044   .0921142     -.1246449    .2367329
   (4 vs 1)  |   .1000573   .0932533     -.0828659    .2829805
--------------------------------------------------------------

.         estimates store american3

. reg refugee i.panel if country==2, robust

Linear regression                               Number of obs     =      1,478
                                                F(3, 1474)        =       2.19
                                                Prob > F          =     0.0869
                                                R-squared         =     0.0044
                                                Root MSE          =     1.2793

------------------------------------------------------------------------------
             |               Robust
     refugee |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .0964033   .0957129     1.01   0.314    -.0913447    .2841512
          3  |   .1578717   .0918334     1.72   0.086    -.0222664    .3380098
          4  |    .233224   .0943895     2.47   0.014     .0480719    .4183761
             |
       _cons |   3.339426   .0663423    50.34   0.000      3.20929    3.469561
------------------------------------------------------------------------------

.         margins r.panel, vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        1.01     0.3140
   (3 vs 1)  |          1        2.96     0.0858
   (4 vs 1)  |          1        6.11     0.0136
      Joint  |          3        2.19     0.0869
             |
 Denominator |       1474
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .0964033   .0957129     -.0913447    .2841512
   (3 vs 1)  |   .1578717   .0918334     -.0222664    .3380098
   (4 vs 1)  |    .233224   .0943895      .0480719    .4183761
--------------------------------------------------------------

.         estimates store refugee3

. reg fair i.panel if country==2, robust

Linear regression                               Number of obs     =      1,478
                                                F(3, 1474)        =       9.08
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0174
                                                Root MSE          =     1.1329

------------------------------------------------------------------------------
             |               Robust
        fair |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .1515477   .0835828     1.81   0.070    -.0124062    .3155016
          3  |   .1915391   .0845476     2.27   0.024     .0256927    .3573854
          4  |   .4231476   .0824746     5.13   0.000     .2613675    .5849277
             |
       _cons |    3.29765   .0595091    55.41   0.000     3.180919    3.414382
------------------------------------------------------------------------------

.         margins r.panel, vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        3.29     0.0700
   (3 vs 1)  |          1        5.13     0.0236
   (4 vs 1)  |          1       26.32     0.0000
      Joint  |          3        9.08     0.0000
             |
 Denominator |       1474
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .1515477   .0835828     -.0124062    .3155016
   (3 vs 1)  |   .1915391   .0845476      .0256927    .3573854
   (4 vs 1)  |   .4231476   .0824746      .2613675    .5849277
--------------------------------------------------------------

.         estimates store fair3

. reg trust i.panel if country==2, robust

Linear regression                               Number of obs     =      1,478
                                                F(3, 1474)        =       4.92
                                                Prob > F          =     0.0021
                                                R-squared         =     0.0097
                                                Root MSE          =     1.2319

------------------------------------------------------------------------------
             |               Robust
       trust |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
       panel |
          2  |   .1229248   .0907541     1.35   0.176    -.0550961    .3009458
          3  |   .1563545   .0894255     1.75   0.081    -.0190602    .3317693
          4  |   .3419324   .0899231     3.80   0.000     .1655415    .5183233
             |
       _cons |   3.219321    .062834    51.24   0.000     3.096068    3.342575
------------------------------------------------------------------------------

.         margins r.panel, vsquish post

Contrasts of adjusted predictions
Model VCE    : Robust

Expression   : Linear prediction, predict()

------------------------------------------------
             |         df           F        P>F
-------------+----------------------------------
       panel |
   (2 vs 1)  |          1        1.83     0.1758
   (3 vs 1)  |          1        3.06     0.0806
   (4 vs 1)  |          1       14.46     0.0001
      Joint  |          3        4.92     0.0021
             |
 Denominator |       1474
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
       panel |
   (2 vs 1)  |   .1229248   .0907541     -.0550961    .3009458
   (3 vs 1)  |   .1563545   .0894255     -.0190602    .3317693
   (4 vs 1)  |   .3419324   .0899231      .1655415    .5183233
--------------------------------------------------------------

.         estimates store trust3          

. 
. // generate figure for diversity and legitimacy outcome, by sample
. coefplot (fair1, color(538g) ciopts(lcolor(538g) recast(rcap))) ///
>                  (fair2, color(538b) ciopts(lcolor(538b) recast(rcap))) ///
>                  (fair3, color(538r) ciopts(lcolor(538r) recast(rcap))), ///
>                  bylabel("(a) UNHCR Is Fair") || ///
>                  (trust1, color(538g) ciopts(lcolor(538g) recast(rcap))) ///
>                  (trust2, color(538b) ciopts(lcolor(538b) recast(rcap))) ///
>                  (trust3, color(538r) ciopts(lcolor(538r) recast(rcap))), ///
>                   bylabel("(b) UNHCR Can Be Trusted") || ///
>                  (american1, color(538g) ciopts(lcolor(538g) recast(rcap))) ///
>                  (american2, color(538b) ciopts(lcolor(538b) recast(rcap))) ///
>                  (american3, color(538r) ciopts(lcolor(538r) recast(rcap))), ///
>                  bylabel("(c) Outcome Good for Americans") || ///
>                  (refugee1, color(538g) ciopts(lcolor(538g) recast(rcap))) ///
>                  (refugee2, color(538b) ciopts(lcolor(538b) recast(rcap))) ///
>                  (refugee3, color(538r) ciopts(lcolor(538r) recast(rcap))), ///
>                  bylabel("(d) Outcome Good for Refugees") vertical ciopts(recast(rcap)) ///
>                  ytitle("Marginal Difference of White-Male Panel vs. Diverse Panels") ///
>                  xtitle("Racial and Gender Composition of Panel (All-White, All-Male as Reference G
> roup)") ///
>                  xlabel(1 `""Mixed-Race" "All-Male""' 2 `""All-White" "Mixed-Gender""' ///
>                                 3 `""Mixed-Race" "Mixed-Gender""', labcol(black)) ///
>                  legend(order(2 "2019 Sample" 4 "2021 Sample (No Country Labels)" ///
>                             6 "2021 Sample (Country Labels)") rows(1)) yline(0) 

.                                 
.                                 
. 
. 
end of do-file

. exit, clear
